/** 
* @file eccv2011-15scenes-fasthik.cpp
* @brief ECCV 2012 Experiment with 15 Scenes
* @author Erik Rodner
* @date 01/17/2012

*/

#include <vector>

#include <core/basics/vectorio.h>
#include <core/basics/Timer.h>
#include <core/basics/Config.h>

#include <vislearning/classifier/fpclassifier/gphik/FPCGPHIK.h>
#include <vislearning/cbaselib/MultiDataset.h>
#include <vislearning/baselib/Globals.h>

#include <gp-hik-core/FastMinKernel.h>
#include <gp-hik-core/FMKGPHyperparameterOptimization.h>
#include <gp-hik-core/parameterizedFunctions/PFAbsExp.h>
#include <gp-hik-core/parameterizedFunctions/PFWeightedDim.h>
#include <gp-hik-core/parameterizedFunctions/PFExp.h>
#include <gp-hik-core/tools.h>



using namespace std;
using namespace NICE;
using namespace OBJREC;

#include "datatools.h"


// this function is very much adapted to Wus files
void readSparseExamples ( const string & fn, Examples & examples )
{
  cerr << "Reading " << fn << endl;
  ifstream ifs ( fn.c_str(), ios::in );
  if ( ! ifs.good() ) {
    fthrow(Exception, "Unable to read " << fn );
  }
  while ( !ifs.eof() )
  {
    int classno;
    if ( !(ifs >> classno) ) break;
    classno--;
    SparseVector *v = new SparseVector; 
    v->restore ( ifs, SparseVector::FORMAT_INDEX_LINE );
    v->setDim( 10000 );
    // evil HACK FIXME
    v->multiply(1.0 / (128.0*1000.0));


    Example example;
    example.svec = v;

    cerr << "read example of size " << v->size() << " belonging to class " << classno << endl;
    cerr << "min: " << v->min() << " max:" << v->max() << endl;
    
    examples.push_back ( pair<int, Example> ( classno, example ) );
  }
  ifs.close();
}

/** 
    
    ECCV 2012 Experiment with 15 Scenes 
    
*/
int main (int argc, char **argv)
{   
  std::set_terminate(__gnu_cxx::__verbose_terminate_handler);

  Config conf ( argc, argv );
 
  FPCGPHIK classifier ( &conf, "GPHIKClassifier" );

  string trainFilename = conf.gS("main", "train"); 
  
  Examples trainExamples;
  readSparseExamples ( trainFilename, trainExamples );

  FeaturePool fp; // ignored anyway
  classifier.train ( fp, trainExamples );

  cerr << "Reading test data file" << endl;
  string testFilename = conf.gS("main", "test"); 
  Examples testExamples;
  readSparseExamples ( testFilename, testExamples );

 
  Timer t;
  Matrix confusion ( 15, 15, 0.0 );
  
  int index = 1;
  for ( Examples::iterator example_i = testExamples.begin(); example_i != testExamples.end(); example_i++,index++ )
  {
    uint classno_groundtruth = example_i->first;

    t.start();
    ClassificationResult r = classifier.classify ( example_i->second );
    uint classno_estimated = r.classno;
    t.stop();
   
    r.scores.store(cerr);
    cerr << "[" << index << " / " << testExamples.size() << "] " << classno_estimated << " " << classno_groundtruth << " time: " << t.getLast() << endl;

    //confusion( classno_groundtruth, classno_estimated ) += 1;
    confusion( classno_estimated, classno_groundtruth ) += 1;
  }

  confusion.normalizeColumnsL1();
  cerr << confusion << endl;

  cerr << "average recognition rate: " << confusion.trace()/confusion.rows() << endl;
  return 0;
}